On Sat, Sep 21, 2019 at 4:10 PM Paul Burton <[email protected]> wrote:
>
> Here are the main MIPS changes for v5.4; please pull.
Hmm. I pulled and because initial tests didn't show any issues, I
already pushed out.
But some unrelated further testing then shows that this:
> Florian Fainelli (2):
> firmware: bcm47xx_nvram: Correct size_t printf format
> firmware: bcm47xx_nvram: Allow COMPILE_TEST
causes problems, and commit feb4eb060c3a ("firmware: bcm47xx_nvram:
Correct size_t printf format") is buggy:
drivers/firmware/broadcom/bcm47xx_nvram.c: In function ‘nvram_init’:
drivers/firmware/broadcom/bcm47xx_nvram.c:151: warning: format ‘%zu’
expects argument of type ‘size_t’, but argument 2 has type ‘u32’ {aka
‘unsigned int’} [-Wformat=]
and the change to use %zu was completely wrong.
It prints out 'header.len', which is an u32, not nvram_len which is a size_t.
Tssk tssk.
I've fixed it in my tree, but this should have shown up in linux-next,
or in MIPS testing. The process clearly failed.
Linus
